(reposted from general discussions forum)
What's the best way to provide loose matching based on color words from free text descriptions?
The idea is to be able to score two different animal descriptions where the description is free text.
As an example, if one person describes a dog as chocolate and cream and the other describes a dog as wheat and brown.
Can Power AI be trained to recognise a heap of different color adjectives as brown, and a different heap as yellow, so that both of the above descriptions are identified as brown and yellow?
Would I be better off doing this parsing the descriptions with regex functions?